1. Peter Richardson IPPP, Durham University
Herwig++
S. Gieseke, D. Grellscheid, K. Hamilton, A. Ribon, PR, P. Stephens, M.H. Seymour, B.R. Webber
M. Baehr, M. Gigg, S. Latunde-Dada, S. Plaetzer, A. Sherstnev, J. Tully
Peter Richardson
IPPP, Durham University
Generartor Services Meeting 22nd Nov

2. Generartor Services Meeting 22nd Nov
Introduction
History
Recent Developments
Current Status
Plans for the Future
Generartor Services Meeting 22nd Nov

3. Generartor Services Meeting 22nd Nov
Introduction
So the FORTRAN HERWIG program has been very successful over the last 20 years so why start again?
Modern computing in both experimental particle physics and industry has moved to object-oriented programming in C++.
Fewer PhD students understand any FORTRAN.
The code structure prevented many physics improvements we wanted/needed to make, e.g.
Showering from SUSY particles
New approaches to matrix element matching
The multi-scale shower.
After 20 years of modifications and improvements the code had become impossible to maintain.
Generartor Services Meeting 22nd Nov

4. Generartor Services Meeting 22nd Nov
Introduction
The number of lines had grown by a factor of over 30.
While the length of the code is not a problem per se the structure was designed with less than 5000 lines of code in mind not
Generartor Services Meeting 22nd Nov

5. Generartor Services Meeting 22nd Nov
Introduction
So a decision was made in 2000 to write a new program in C++ based on the same physics philosophy but incorporating new theoretical developments.
There was a first release including the shower and hadronization for e+e- collisions.
Herwig++ 1.0, S. Gieseke, A. Ribon, M. H. Seymour, P. Stephens, B.R. Webber JHEP 0402:005, 2004.
Generartor Services Meeting 22nd Nov

6. Generartor Services Meeting 22nd Nov
Shower Improvements
Smooth coverage of the soft region.
Better evolution for heavy particles.
No overlapping regions of phase space.
Generartor Services Meeting 22nd Nov

7. Generartor Services Meeting 22nd Nov
Shower Improvements
The main aim was to allow evolution down to zero pT for radiation from massive particles and to avoid the ‘dead-cone’ approximation used in the FORTRAN program.
Generartor Services Meeting 22nd Nov

8. Generartor Services Meeting 22nd Nov
Recent Developments
A new version of Herwig++, 2.0, was released on the 29th of September,
hep-ph/ , S. Gieseke, D. Grellscheid, K. Hamilton, A. Ribon, P. Richardson, M.H. Seymour, P. Stephens, B.R. Webber.
This version includes many major developments:
Simulation of initial-state radiation;
Simulation of the underlying event;
Many more hadron-hadron matrix elements;
Simulation of initial-state radiation in the decay of heavy particles.
Generartor Services Meeting 22nd Nov

9. Generartor Services Meeting 22nd Nov
Recent Developments
Matrix element correction for Drell-Yan and top decays;
Simulation of QED radiation in particle decays.
This is the first version which can be used for realistic physics studies of hadron-hadron collisions.
Generartor Services Meeting 22nd Nov

10. Generartor Services Meeting 22nd Nov
New Matrix Elements
The matrix elements for:
Direct photon pair production;
Photon+jet production;
Higgs boson production;
Higgs boson plus jet production;
Heavy quark pair production;
QCD 2g2 scattering processes;
were included.
Other processes are available using a Les Houches accord interface.
Generartor Services Meeting 22nd Nov

11. Initial-State Radiation
The major changes in this version were
Simulation of initial-state radiation for processes other than Drell-Yan;
Final-State radiation from the partons produced in the initial-state shower;
The hard matrix element correction for Drell-Yan W and Z production.
Generartor Services Meeting 22nd Nov

12. pT of the Z compared with CDF data
Generartor Services Meeting 22nd Nov

13. Generartor Services Meeting 22nd Nov
Top Decay Shower
A key feature of the new parton shower algorithm is the simulation of radiation in particle decays.
In the new approach the decaying particle emits radiation in its decay to ensure the full soft region is covered.
However the soft matrix element is required to give smooth coverage in the soft region.
Generartor Services Meeting 22nd Nov

14. Top Shower and Matrix Element Correction in tgbW+g
hep-ph/ , Keith Hamilton, Peter Richardson.
Generartor Services Meeting 22nd Nov

15. Generartor Services Meeting 22nd Nov
Underlying Event
Uses the UA5 model as in the FORTRAN program to gives some simulation of the underlying event.
Generartor Services Meeting 22nd Nov

16. Simulation of QED Radiation
The FORTRAN program did not simulate QED radiation.
Recently this was included in the C++ program using the YFS formalism.
K. Hamilton and PR hep-ph/ , JHEP 0607:010, 2006.
Generartor Services Meeting 22nd Nov

17. Bug tracking and Release Strategy
We are using HEPFORGE for bug tracking and a wiki
Our plan is to release a monthly update which will only contain bug fixes etc.
There will be less frequent major releases with physics changes to the simulation.
Generartor Services Meeting 22nd Nov

18. Generartor Services Meeting 22nd Nov
Plans for the Future
The 2.0 release gives us a solid foundation on which further developments can build.
There are a number of features which we still need to include:
JIMMY multiple scattering model for the underlying event;
different kinematic reconstruction procedures for the shower;
BSM Physics;
new hadron decay model;
spin correlations throughout the simulation.
In order for the simulation to be as good as, or better than, the FORTRAN for everything.
There has been a lot of work on the hadron decays and BSM physics.
Generartor Services Meeting 22nd Nov

19. Generartor Services Meeting 22nd Nov
BSM physics
The new hard processes are based on a library of classes using the helicity formalism.
For other BSM models we will automatically generate the decays and hard processes from the Feynman rules (Vertices).
Using helicity amplitudes throughout will make it easy to add spin correlations.
My PhD student Martyn Gigg is doing most of the work.
Generartor Services Meeting 22nd Nov

20. Generartor Services Meeting 22nd Nov
Example SUSY Process
Generartor Services Meeting 22nd Nov

21. Generartor Services Meeting 22nd Nov
Hadron Decays
We have a wide range of hadron decay matrix elements available, including spin correlations.
Much of the code is ready it just needs to be full tested.
Hopefully this will happen in the near future.
Generartor Services Meeting 22nd Nov

22. Generartor Services Meeting 22nd Nov
Hadron Decays
The simulation includes detailed modelling of many decays.
Generartor Services Meeting 22nd Nov

23. Generartor Services Meeting 22nd Nov
Hadron Decays
Spin correlations are also included.
In the decay the angle between the tau decay planes, f*, and between the pions, d*, depends on whether the parity of the Higgs boson.
Generartor Services Meeting 22nd Nov

24. Future Shower Improvements
In addition to the other features one of the main reasons for going to C++ was to allow improvements to the shower algorithm:
CKKW matrix element matching;
The multi-scale shower;
The Nason approach to
Generartor Services Meeting 22nd Nov

25. Generartor Services Meeting 22nd Nov

26. Nason Approach to MC@NLO
hep-ph/ Oluseyi Latunde-Dada, Stefan Gieseke, Bryan Webber
Generartor Services Meeting 22nd Nov

27. Generartor Services Meeting 22nd Nov
Summary
We have made a lot of progress in the last year.
Herwig++ is now ready for hadron collisions.
Following this first release for hadron collisions further improvements will now be easier and happen faster.
Generartor Services Meeting 22nd Nov